IOT and Indoor Localization

IOT and Indoor Localization
Dr. David Chieng
Wireless Innovation Lab
MIMOS, Berhad
Malaysia
Content
•
•
•
•
•
•
•
•
IoT
Location, the missing context?
Motivations for getting indoor
Wireless indoor positioning techniques
Deployment approaches
MIMOS Indoor Location Platform
Research challenges & potential solutions
Conclusions
2
ASEAN RISE 2016, Hanoi
IoT
• Billions of devices  around us
• Billions worth of market opportunities?
Wireless
Sensor
Networks
Access
Networks
Operations
Management
Applications/
Services
Markets and Markets, Nov 2014
• Smart home, smart office, smart health, smart
manufacturing, smart retail, etc  Indoor
3
ASEAN RISE 2016, Hanoi
Location, the Missing Context?
• Intelligent = Context-aware
• 5 elements of context: Who, What, Why, When
and WHERE
• Typical context-aware IoT applications. E.g.
– Play my favourite music(what) when I enter my(who)
bedroom (where)
– Call nearest(where) person(who), when home
alarm(what) triggered
– When did Johnny(who) reach/left school(where)?
• In indoor environment, the “where” is largely
missing
– 80% people are indoor, 80% of the time….
4
ASEAN RISE 2016, Hanoi
What can Location Info offer for
IoT?
• With location awareness, a more meaningful interactions
between human, things, events and location can take
place
• Semantic positioning – beyond geo spatial info.
Deriving user’s position & action through IoT sensing
• Such a rich set of contextual info can be translated to a
wide range of innovative location-based applications:
– Trigger services based on what user is doing?
– Advertise based on user’s state?
5
ASEAN RISE 2016, Hanoi
Motivations for Getting Indoor
• Buildings getting higher, shopping malls getting
bigger
• “ABI Research forecasted that total indoor
location revenues will reach US$10 billion in
2020, driven primarily by BLE Beacons and
advertising”, May 2015.
• Close to 50 shopping malls in Klang Valley alone
and around 10 more to be added by end of this
year.
• Stiff competition implies the need to differentiate
6
ASEAN RISE 2016, Hanoi
Shopping Malls in Klang Valley
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Mid Valley Megamall
Capital Square
Sogo Kuala Lumpur
Suria KLCC
Ampang Park
Intermark
Avenue K
Pavilion Kuala Lumpur
Fahrenheit 88
Lot 10
Low Yat Plaza
Starhill Gallery
Sungei Wang Plaza
Viva Home
Leisure Mall
Sentral Mall (Cheras)
Kenanga Wholesale City
Berjaya Times Square
Quill Mall
Nu Sentral
Sunway Putra Mall
Festival Mall (Setapak)
1 Utama Shopping Center
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Sunway Pyramid
Jaya Shopping Centre
Subang Parade
Empire Subang
Citta Mall
Center Point
IOI Mall (Puchong)
IOI City Mall (Putrajaya)
Alamanda (Putrajaya)
Tropicana City Mall
Ikano (Power Station)
The Curve (eCurve)
Publika
Setia City Mall
Paradigm Mall
Mines Resort City
Shaw Centrepoint (Klang)
Klang Parade
AEON Bukit Tinggi Shopping Centre
One City Mall
Gateway (KLIA2)
Mitsu Outlet
At least 10 more malls to be
opened in 2016
ASEAN RISE 2016, Hanoi
7
Wireless Positioning Techniques
• Trilateration (TOA, TDOA, RSSI strength)
• Triangulation (angle-based)
• Fingerprinting (pattern-based)
– Zero or minimal infra cost
– Fast setup
– Device availability  smartphones
– Suited for Indoor
8
ASEAN RISE 2016, Hanoi
Deployment Approaches
• Two main approaches:
1. Infrastructure dependent
a) Green field - need large scale deployment of
devices (WiFi/BLE/Femto/Light/Sound)
b) Brown field – only require to install an app in
smart phone (relying on EXISTING APs or
BLEs).
2. Infrastructure less – based on built-in
sensors such as magnetometer, gyroscope,
accelerometer, etc.
9
ASEAN RISE 2016, Hanoi
MIMOS Indoor Location Platform
• Use existing WiFi or BLE
signals
• Smartphone-based
(fingerprinting)
• Simple and intuitive setup
process
• Modified Bayesian
estimation
• Accuracy within ~ 5 to
10m
10
ASEAN RISE 2016, Hanoi
Mi-Loc: Software Architecture
User
Applications
APIs
11
ASEAN RISE 2016, Hanoi
Potential Services
Panic Button
12
ASEAN RISE 2016, Hanoi
Pilot Trial in IOI City Mall
13
ASEAN RISE 2016, Hanoi
Research Challenges & Solutions
• Within fingerprinting approach:
– Device heterogeneity. Potential solutions:
• Relative signals
• Pattern-based
– Dynamic wireless environment. Potential
solutions:
• Crowdsensing/data collection
• Semi permanent calibrator
• There is a need to have hybrid
approaches - integrating with sensorbased tracking e.g. step sensor
14
ASEAN RISE 2016, Hanoi
Mean Error (m)
Device Heterogeneity study in real
environment (mall data)
15
ASEAN RISE 2016, Hanoi
Conclusions
• Location is a critically missing context for
IoT applications/services indoor.
• With location info, a richer variety of new
applications/services can be created with
pervasive interactions with networked of
things.
• More interesting with sub meter
granularity.
• Lots of challenges but it is getting better
16
ASEAN RISE 2016, Hanoi
17
ASEAN RISE 2016, Hanoi